A Study on Language Cognitive Mechanisms from the Perspective of the Integration of Psycholinguistics and AI

This paper explores the cognitive mechanisms of human language understanding and production from the perspective of psycholinguistics-AI integration. It reviews traditional psycholinguistic models and real-time processing mechanisms of language processing, and combines them with the technical evolution of AI language modeling, especially the development of large-scale pretrained language models driven by deep learning. Then, this paper proposes a Distributed-Symbolic Hybrid Model, which is a theoretical framework of language cognition that integrates distributed semantic representation and symbolic structure reasoning. This model is theoretically compatible with psycholinguistic evidence and AI language modeling results. By simulating the functional characteristics of semantic cortex areas (such as the TPJ) and syntactic processing related brain areas (such as Broca’s area) in the human brain, the model enhances biological plausibility. The paper aims to build a language cognitive model with dynamic causal mechanisms, provide AI with interpretable cognitive constraints, and offer computational implementation tools for psycholinguistic theories.